AI Engineer – Mid Level
Kathmandu, Nepal (Hybrid)
About the Role
As a Mid-Level AI Engineer at Dogma Group, you will work in a client services environment delivering AI solutions under real deadlines and changing requirements. This is a hands-on role centered on building agentic systems backed by LLMs and computer vision applications, then deploying them as containerized, production-ready services. You will work with multi-agent architectures, modern retrieval pipelines, and standardized agent protocols, translating client requirements into working solutions within compressed timelines.
Key Responsibilities
- Design, build, and deploy single-agent and multi-agent systems backed by LLMs, using LangChain and LangGraph where appropriate, or built from scratch in Python with Pydantic for structured outputs and data validation
- Connect agents to tools, data sources, and other agents using standardized protocols such as MCP (Model Context Protocol) and A2A (Agent-to-Agent)
- Build production-grade RAG pipelines using hybrid retrieval (dense vector search combined with keyword search) and reranking with cross-encoders to improve answer grounding and reduce hallucination
- Work with vector databases such as Pinecone, Weaviate, Qdrant, or pgvector, selecting the right retrieval architecture for each client's scale and data requirements
- Build and integrate computer vision solutions including object detection, image classification, segmentation, and OCR, using OpenCV, TensorFlow or PyTorch, and vision-language models for multi-modal understanding
- Containerize AI services using Docker and manage deployments across cloud and on-premise environments
- Deploy and maintain models and agentic services in production, with observability, tracing, and evaluation pipelines to monitor performance, drift, and reliability
- Translate client requirements into technical scope and deliver iteratively under tight timelines
- Build data pipelines for training and inference, ensuring clean and reliable data flow
- Write maintainable Python code with proper documentation, testing, and version control practices
- Communicate progress and technical tradeoffs clearly to both technical and non-technical stakeholders
- Participate in client calls, demos, and reviews as a technical representative of the team
- Evaluate new AI tools, models, protocols, and frameworks for adoption across client projects
Experience/Skills Required:
- 3+ years of relevant work experience in AI/ML development and deployment, with a Bachelor's Degree in Computer Science, Computer Engineering, Data Science, Machine Learning, or a related field
- Foundational knowledge of machine learning and deep learning concepts, including model training, evaluation, and optimization
- Mathematical knowledge in statistics, probability, linear algebra, and calculus as applied to ML and DL
- Strong knowledge of GenAI and LLMs, with hands-on experience building and integrating LLM-based applications
- Hands-on experience developing agentic systems in Python, building agents from scratch using Pydantic for data validation
- Familiarity with agentic frameworks such as LangChain and LangGraph, and standardized agent protocols such as MCP and A2A
- Experience with vector databases such as Pinecone, Weaviate, or Qdrant,, and an understanding of hybrid search and reranking for retrieval-augmented generation
- Proficient in TensorFlow and PyTorch, with a track record of taking models from development through deployment
- Experience with computer vision libraries and tools including OpenCV, YOLO, PIL, and deep learning frameworks for image processing tasks
- Hands-on experience containerizing and deploying AI services using Docker, with working knowledge of orchestration tools
- Working knowledge of at least one major cloud platform (Azure preferred) for AI/ML deployment
- Strong analytical thinking, problem-solving ability, and clear communication for both technical and non-technical stakeholders
- Comfort working in a fast-paced, client-facing environment with shifting priorities and short delivery cycles.
- Awareness of privacy, data security, and ethical considerations in GenAI development, including responsible handling of client data and compliance with relevant regulations.
We respect your privacy and do not share your data with any third party by handling them responsibly in compliance with the laws. For more information, please read our Privacy Policy and Terms and Conditions.
